Xiao Liu MBChB PhD
I am an Associate Professor in AI and digital health technologies at the University of Birmingham. I am interested in the translation of scientific evidence into best practice across research, policy and regulation. Previously, I was a Health Scientist at Apple, and before that I was an ophthalmologist in the UK's National Health System.Alastair Denniston and I co-lead the AI and Digital Health Policy and Research Group at University Hospitals Birmingham/University of Birmingham, where we focus on three specific aspects of AI in health:
1) Safety
2) Effectiveness
3) EquityMy primary motivation is to bring the best digital technologies to patients to improve their lives and our understanding of health and disease. To do so, I believe we need to co-create an efficient, innovative and responsible ecosystem with seamless cross-sectoral collaboration. I believe policy should be rooted in evidence. I hate waste and duplication, and I love creativity and pragmatism.
About me
Hi, I'm Xiao (Xiaoxuan) Liu. Xiao is pronounced "Shau".I am an ophthalmology doctor and a researcher at University Hospitals Birmingham NHS Foundation Trust and The University of Birmingham.I co-led the SPIRIT-AI & CONSORT-AI initiative in 2020: reporting standards for clinical trials evaluating AI systems - EQUATOR Network endorsed guidelines, which have since been adopted by international medical device regulators and medical journals including Nature Medicine and the Lancet. I am also on the steering group for DECIDE-AI, TRIPOD+AI and STARD-AI, reporting guidelines for early feasibility studies, prediction and prognostic models and diagnostic accuracy studies for AI, respectively.Recently, I led STANDING Together - a project on tackling bias in health datasets to ensure AI benefits all. STANDING Together provides guidance for documentation of datasets to promote transparency, as well as guidance on how to address these biases and mitigate risks to minorities
and underserved groups in the context of AI as a Medical Device. STANDING Together was referenced at the UK Safety Summit in 2023, and at the G20 Summit in Brazil in 2024.I work with government and policy institutions on their approach to evaluating AI in healthcare, including the WHO, MHRA, NICE, NHS England, BSI and the NHS AI Lab.I am a Deputy Editor at NEJM-AI. A journal focused on identifying and evaluating state-of-the-art applications of artificial intelligence to clinical medicine.Prior to this, I completed my PhD on automated imaging-based methods for measuring inflammation in the eye, under the supervision of Alastair Denniston, Pearse Keane and David Moore.My work has been featured in Wired, The Guardian, BBC Radio, The New Scientist and other news outlets, and recognised as Top Notable Advances of 2019 and 2020 by Nature Medicine.
Talks
Upcoming:Australian Academy of Health and Medical Sciences Annual Meeting, October 2024Singapore National AI Conference, December 2024------
Previous:SAIL 2023, 9-12th May 2023 - Panel: AI in Clinical TrialsWHO/ITU Focus Group for AI in Health conference - 21-24th March 2023, MIT/Harvard Cambridge - Workshop on “Deployment of AI technology in real-world settings”February 2023: British Institute of Radiology Women in Imaging - "First do no harm: a responsible approach to AI in health".June 2022: ACM FAccT CRAFT Session Panel on Communication Across Communities in Machine Learning Research and Practice.April 2022: Intelligent Health UK - STANDING together - "developing standards for datasets underpinning AI systems so they are diverse, inclusive and can work across all demographic groups".
Community
The Alan Turing Institute
Special Interest Group for Clinical AI - Co-lead (2022-23)Data Science for Health Equity
Standards for Health Data Equity - Theme Lead (2022-23)
Selected Publications
Full list of publications here.The value of standards for health datasets in artificial intelligence-based applications Arora M, Alderman JE, Palmer, J, Ganapathi S, Laws E, ...Liu X. Nature Medicine 2023.What's fair is… fair? Presenting JustEFAB, an ethical framework for operationalizing medical ethics and social justice in the integration of clinical machine learning: JustEFAB McCradden M, Odusi O, Joshi S, Akron I, et al. FAccT 2023.Tackling bias in AI health datasets through the STANDING Together initiative
Ganapathi S, Palmer J, Alderman JE, Calvert M, Espinoza C, Gath J, Ghassemi M, Heller K, Mckay F, Karthikesalingam A, Kuku S, Mackintosh M, Manohar S, Mateen BA, Matin R, McCradden M, Oakden-Rayner L, Ordish J, Pearson R, Pfohl SR, Rostamzadeh N, Sapey E, Sebire N, Sounderajah V, Summers C, Treanor D, Denniston AK & Liu X
Nature Medicine 2022.Editorial: Surfacing best practices for AI software development and integration in healthcare Sendak M, Vidal D, Trujillo S, Singh K, Liu X and Balu S. Front. Digit. Health 2023.The Medical Algorithmic Audit
Liu X, Glocker B, McCradden MM, Ghassemi M, Denniston AK, Oakden-Rayner L.
The Lancet Digital Health 2022.Reporting guidelines for clinical trial reports for interventions involving artificial intelligence: the CONSORT-AI extension
Liu X, Rivera SC, Moher D, Calvert MJ, Denniston AK & The SPIRIT-AI and CONSORT-AI Working Group.
Nature Medicine 2020.Guidelines for clinical trial protocols for interventions involving artificial intelligence: the SPIRIT-AI extension
Rivera SC, Liu X, Chan A-W, Denniston AK, Calvert MJ & The SPIRIT-AI and CONSORT-AI Working Group.
Nature Medicine 2020.A comparison of deep learning performance against health-care professionals in detecting diseases from medical imaging: a systematic review and meta-analysis
Liu X, Faes L, Kale AU, Wagner SK, Fu DJ, Bruynseels A, Mahendiran T, Moraes G, Shamdas M, Kern C, Ledsam JR, Schmid MK, Balaskas K, Topol EJ, Bachmann LM, Keane PA, Denniston AK.
The Lancet Digital Health 2019.Health data poverty: an assailable barrier to equitable digital health care
Ibrahim H, Liu X, Zariffa N, Morris AD, Denniston AK.
The Lancet Digital Health 2021.A global review of publicly available datasets for ophthalmological imaging: barriers to access, usability, and generalisability
Khan SM, Liu X, Nath S, Korot E, Faes L, Wagner SK, Keane PA, Sebire NJ, Burton MJ, Denniston AK.
The Lancet Digital Health 2021.Characteristics of publicly available skin cancer image datasets: a systematic review
Wen D, Khan SM, Ji Xu A, Ibrahim H, Smith L, Caballero J, Zepeda L, de Blas Perez C, Denniston AK, Liu X, Matin RN.
The Lancet Digital Health 2022.